Based on the hypothesis that rare alleles are in mutation and random loss equilibrium, mutation rate can be indirectly estimated by measuring the number of rare variants and the average existing time of a mutant allele. This method can be applied to estimate the mutation rate in humans. However, this estimation of mutation rate is affected by the presence of premeiotic clusters of mutation. Mutation clusters change both the number of initial mutants and the average existing time of a mutant allele. As a result, the formula indirectly estimating mutation rate should be modified. The influence of premeiotic clusters is more obvious when the population size is small or the average cluster size is big. For example, if the population size is 3,000 and average cluster size is two, instead of one, the mutation rate is increased by about 9.4%.